| Summary: | 碩士 === 國立高雄第一科技大學 === 金融理財研究所 === 98 === This study tried to examine whether VIX can forecast the realized volatility of TAIEX. Because of the non-stationarity of both VIX and realized volatilities, we calculated the change rate of the two series to test their lead-lag relationship by VAR and Granger causality tests. We found the lag variables could significantly explain both VIX and realized volatilities in VAR test. On the other hand, if we used Granger Causality test and increased the number of lag progressively to 3, we found the realized volatilities led VIX but the reverse was not true. It seems VIX can’t forecast realized volatility. Reversely, VIX is affected by realized volatility. Instead of being able to forecast future market volatility, this result implies that investors’ prediction is greatly based on realized volatility.
|
|---|
